Welcome back to the Miles Deutscher weekly newsletter!
This week, we're breaking down 1 workflow: an AI that runs your charts, and the 10 prompts that prove it actually works.
AI x Trading builds the connection. Your rules go in once, as a standing filter. The model checks every answer against them.
AI Skills & Prompts is the test. 10 prompts, run in order, that tell you if the connection is real or if it's just agreeing with you.
Markets ran the same test at scale. Bitcoin closed above its 50-week moving average for the first time in 45 weeks. Same week, Ukraine hit Russian energy infrastructure with one of the largest drone attacks of the war. That headline used to guarantee a red day. Not this week.
Inside MHC, the filter had a name: discipline. Fabian skipped the tempting rotation. Morin waited on a level that hasn't broken.
Same rule, every section: don't trust the setup until you've tried to break it.
Start with the connection.
IN TODAY'S READ
Table of Contents
After our main breakdown, stick around for AI Skills & Prompts, this week's Markets catalysts, and the MHC Recap of setups from inside the room.
AI X TRADING: This week's AI workflow built for traders
Wire your charts to a model that never gets tired of screening
A TradingView MCP connection changes the job. The model stops being something you ask about a chart. It starts running the chart.
Hand it your rules once: risk tolerance, position size, invalidation levels, watchlist. From there it does 4 things on command. Pulls a live bull and bear case on any ticker. Screens the market for setups that match your rules, not generic advice. Backtests an idea and flags if it's overfitted. Flags the moment a watchlist level gets hit.
Why it matters: most people use AI for narrative, not signal. This flips that.
The rules go in once, as text. Every answer after that gets filtered through them. That's the real mechanism, not that the model got smarter. Your risk parameters become a standing filter, not something you remember to apply at 2am when a coin spikes.
Same logic on the backtest. Flagging overfitting means re-running the setup across historical windows and checking if the edge survives small changes. Not 1 clean equity curve from the window that happened to work.
The honest limitation: the MCP connection sees price and indicators. It doesn't see your conviction, your sizing history, or what a room full of traders already discarded 3 hours ago.
A bot that flags every setup matching your rules on paper will flag things better traders already ruled out. It has no memory of that conversation.
That's the same gap a scheduled brief always has. It tells you what happened. Not what the room already knew before it happened. [Get the room →]
Before next Tuesday, in under 30 minutes: pick 1 ticker you already watch. Connect the MCP server. Paste in your real rules, real size, real invalidation level. Run the bull/bear prompt once.
Then run /tv-health-check. Confirming the connection didn't silently drop is the most common failure mode here.
Don't screen the whole market on the first run. 1 ticker. Your real rules. A level you already have an opinion on. That's the test.
AI Skills & Prompts: New tools, the mechanism behind them, what to do with it
10 prompts to stress-test your TradingView MCP connection
You ran /tv-health-check previously. Connection confirmed, 1 ticker tested.
This is the rest of the workout. 10 prompts, built to test everything the model can actually do once it's reading your charts, not just answering about them.
Once the connector's live, copy these in, top to bottom.
01. CONNECTION CHECK
"What symbol and timeframe is my current chart on, and what's the latest price?"
02. TECHNICAL SNAPSHOT
"Pull the setup on [TICKER]: key levels, trend, RSI, MACD, moving averages, then give me the bull and bear case."
03. MULTI-TIMEFRAME
"Analyze [TICKER] across 15m, 4H, and 1D. Where do the timeframes agree and disagree?"
04. SCREENER
"Screen the market for bullish RSI divergence on the 4H with volume above the 20-day average."
05. LEVELS & STRUCTURE
"Mark the key support, resistance, and invalidation on [TICKER] and explain why each matters."
06. DRAW THE SETUP
"Draw a labelled rectangle at the [TICKER] demand zone between $X and $Y, then read it back to confirm."
07. WATCHLIST SCAN
"Scan my watchlist and flag anything within 3% of a major level or showing an overnight volume spike."
08. PINE SCRIPT + BACKTEST
"Write my [strategy] as a Pine Script indicator, then show me the backtest and signal frequency."
09. RELATIVE STRENGTH
"Compare [TICKER A] vs [TICKER B] on relative strength over 90 days. Which is leading?"
10. MORNING ROUTINE
"Every morning at 8am, scan my watchlist, flag setups matching my rules, and send a one-page brief."The rule: #01 first, every time. If it doesn't return your real chart and real price, the connection isn't live, everything after that is the model guessing, not reading.
After that, work down in order. #02 through #05 test if it actually understands your chart. #06 through #09 test if it can act on it. #10 is the 1 worth keeping running: your standing morning brief.
Do this: run #01 right now if you haven't already. Then pick 1 more from the list, on 1 real ticker, before you touch the rest.
MARKETS: The catalysts and moves that matter this week
Market update
→ Bitcoin closed the week above its 50-week moving average for the first time in 45 weeks, a level that's historically marked confirmation that bear market lows are in. Bitcoin is up 29% in 35 days.
That reclaim happened in the same week Ukraine ran one of its largest drone attacks of the war, more than 1,600 drones aimed at the Moscow region, hitting the refinery that supplies over a third of the region's fuel. The strike landed 6 days after Trump said Ukraine and Russia had agreed not to hit each other's energy targets.
→ Why that combination matters more than either headline alone: a strike like that used to be a guaranteed red day across risk assets. It wasn't this week. Crypto absorbed a live escalation, on an energy target, right after a diplomatic agreement broke, and kept its technical structure intact.
The sector didn't just hold, it rotated hard into risk. L1s led the 5-day tape at 29%, with DeFi lending and yield, DEX and trading, stablecoins and RWA, and crypto infrastructure all posting 20%+ weeks. That's an L1 infrastructure story, not just a Bitcoin story, and it's happening while traditional defensive sectors, financial advisors, real estate, payments infrastructure, closed the same 5 days red.
→ Positioning, in 3 tiers rather than 1 trade: highest conviction is Bitcoin on the moving-average reclaim itself, since that signal has a real historical base rate behind it and doesn't depend on any single narrative holding. Second tier is the L1 infrastructure basket that led this week, DEX, lending, RWA rails, sized smaller, because a 1-week sector sweep is a rotation, not yet a trend worth full size. Bottom tier, smallest size, is anything catalyst-specific off the watchlist below: a real date or a real mechanism, not a general "alts are ripping" bet.
→ Now the crypto watchlist for the week ahead:
MHC RECAP: Setups and wins from inside Miles High Club
This Week Inside MHC: A Few Setups We're Watching
→ Fabian on $XMR ( ▲ 1.23% )
His original XMR thesis played out, but ZEC "vastly outperformed" it, and he's explicit that the tempting move, rotating into the laggard because it's cheaper, is usually the one that leads to underperformance this early in a rally. He's still holding ZEC as the better mid to long term position, but he separates that from a shorter trade: XMR looks clean for continuation up to the January high near $800, which he's flagged before as a magnet, and he's treating that as a trade idea, not a reason to trim the core ZEC hold.
→ Paradise on $USELESS ( ▲ 13.3% )
No new position yet, just a condition. He’ll add to the existing bag only if price flushes into a specific pattern, equal lows feeding into the daily consolidation she's watching, rather than adding on strength or out of fear of missing the move. The add is conditional on the setup coming to her.
→ Morin on $OP ( ▼ 2.6% )
OP has failed 3 times to clear its range high, and Morin's angle isn't the chart alone, it's that OP tends to follow ARB with a lag, and he flagged ARB in this same channel before its breakout. He's not entering yet. The plan is to get long only on a confirmed break of $0.112, with the swing high at $0.18 as the main target, which means the setup only exists if the level actually breaks, not before.
→ Surfxbt on $PUMP ( ▲ 3.91% )
Got 70% of the intended risk filled at the lows, then trailed stops and manually entered the rest on a 4H diagonal breakout plus a pivot level reclaim. He's managing the position on a D1 trend basis instead of reacting to every 4H print, with the local highs as the first trouble area and the listing pivot high as the real target, and he's applying the same fill discipline to his spot bags.
→ Room wins of the week
Daniel smashed it with a UNI long, which brought him a 483% ROI.
A member posted a +164% long on NEAR off Surfxbt’s call.
Remember: not financial advice; always do your own due diligence.
CONTENT PIPELINE: Latest content from Miles Deutscher Finance
New From Miles
YouTube: [Latest video title]. Watch now →
𝕏: [How to produce beautiful design with GPT-6 Astra (step-by-step).]. Read now →
Skool: New prompts, workflows, research & more. Join here (100% free) →
MHC: Live setups and analysis from our traders, every day. Join MHC →
See you next Tuesday,
Miles Deutscher & team
Want to sponsor or partner with us? Reach out at [email protected]
